Lectures
The course comprises 10 lectures.
| Lecture | Lecturer | Reading | Date | Slides |
Recording and discussion forum |
|
| 1. | Introduction, ML basics, linear models | NW | [GBC] 1, 5, [LWLS] 2.1, 3.1-3.2 | 17/3, 13:15-15:00 | pdf Download pdf | Lecture 1 |
| 2. | Feed forward neural networks | NW | [GBC] 6.1-6.4, [LWLS] 6.1 | 24/3, 13:15-15:00 | pdf Download pdf | Lecture 2 |
| 3. | Optimization: Stochastic gradient and backpropagation | TS | [GBC] 8.1-8.3,6.5, [LWLS] 6.2, 5.4-5.5 | 31/3, 13:15-15:00 | pdf Download pdf | Lecture 3 |
| No lecture, but there will be a helpdesk session 15:00-16:00 | 7/4 | |||||
| 4. | Convolutional neural networks 1 | JL | [GBC] 9, [LWLS] 6.3 | 14/4, 13:15-15:00 | pdf Download pdf | Lecture 4 |
| 5. | Convolutional neural networks 2 | JL | [GBC] 9 | 21/4, 13:15-15:00 | pdf Download pdf | Lecture 5 |
| 6. | Over-/underfitting, bias-variance, regularization | NW | [LWLS] 4, 6.4 [GBC] 7 | 28/4, 13:15-15:00 | pdf Download pdf | Lecture 6 |
| 7. | Practical methodology and batch normalization | NW | [GBC] 8.7.1, 11, [LWLS] 11 | 5/5, 13:15-15:00 | pdf Download pdf | Lecture 7 |
| 8. | Deep time series models 1 | CA | [GBC] 10 | 12/5, 13:15-15:00 | pdf Download pdf | Lecture 8 |
| 9. | Deep time series models 2 | CA | 19/5, 13:15-15:00 | pdf Download pdf | Lecture 9 | |
| No lectures, but there will be a helpdesk sessions 15:00-16:00 | 26/5, 2/6, 9/6 | |||||
| 10. | Project proposal presentation | 16/6, 13:15-15:00 |
After each lecture there will be a helpdesk and Q/A session. 15:00 - 16:00. There is no mandatory presences at the lectures. Only at the project proposal presentations, (if you aim for the optional project extension of the course)
The lectures will be recorded and links will be made available above. You need to be enrolled to the course and logged in to access the recording.
The recommended book for the course is
- [GBC] Ian Goodfellow, Yoshua Bengio and Aaron Courville Deep Learning
, MIT Press, 2016.
We will not follow [GBC] strictly and we do not cover all aspects of the suggested chapters in the lecture. For some lectures (partly 1-4, 7) in the course, the following book also covers material in a more condensed format.
- [LWLS] Andreas Lindholm, Niklas Wahlström, Fredrik Lindsten, and Thomas B. Schön Machine Learning - A First Course for Engineers and Scientists Links to an external site. (previously named Supervised machine learning)
Another great resource is
- [N] Michael A. Nielson Neural Networks and Deep Learning Determiniation Press, 2015.
which is a bit more hands-on in comparison to [GBC] but does not cover as much and is lacking some details.
For lecture 8 and 9 some additional resources will be used.
NW = Niklas Wahlström
TS = Thomas Schön
JL = Joakim Lindblad
CA = Carl Andersson